The new game of operations with AI
AI crashes the price of the mechanical part of operations: routing, scheduling, the first read of the dashboard. What starts being worth something is your judgment about root cause and process redesign. And there's a catch that's deadly in operations: automating a broken process just speeds up the broken part.
Monday morning, six thirty, and the night shift left three open incidents: an order that didn't go out on time, a route that came back half done, and a customer who called in complaining twice. Before, that used to be your entire first hour: pull the dashboard, cross-reference the shift spreadsheet, put together the summary to pass to the team. Today AI reads the system and hands you that summary in seconds. The question that separates the expensive professional from the cheap one is no longer "do you know how to build the shift summary?". It's: "do you know why these three things happened, and what changes so they don't happen again?".
Monthly logistics cost close, and AI already delivers, in minutes, the cost variance by route, something that used to take a whole afternoon cross-referencing spreadsheets. But it also points, with the same confidence as always, to fuel as the source of the increase, when it actually came from overtime not yet logged in the system. What separates the expensive professional from the cheap one is no longer building the cost spreadsheet: it's knowing which explanation matches the real operation.
A carrier contract has a late-delivery penalty clause that AI summarizes in seconds, together with the SLA promised to the end customer. But it also assumes, with the same confidence as always, that the two SLAs are equivalent, when the internal contract is actually stricter than what's promised to the customer. What separates the expensive legal professional from the cheap one is no longer reading the contract: it's knowing where the fine print actually squeezes the operation.
This quarter's campaign promises 24 hour delivery, and AI already drafts the ad in minutes from the brief. But it also assumes, with the same confidence as always, that the operation can sustain that promise every day, when in practice it only holds up outside peak times. What separates the expensive marketing professional from the cheap one is no longer writing the ad: it's knowing whether the operation can back up the promise before you publish it.
Next month's schedule ready in minutes, AI already suggests how many people per shift based on history. But it also assumes, with the same confidence as always, an average productivity per shift that ignores the fact that the overnight shift runs with half the team's experience. What separates the expensive operational HR professional from the cheap one is no longer building the schedule: it's knowing which shift really needs reinforcement.
The delivery app's roadmap prioritizes real time tracking, and AI already builds the PRD in minutes from the support tickets. But it also assumes, with the same air of certainty, that tracking solves the most common complaint, without having cross-checked that the real complaint is about delivery time, not visibility. What separates the expensive PM from the cheap one is no longer documenting: it's knowing which problem the operation actually feels.
A big account asks if you can handle double the volume during the year end peak, and AI already drafts the answer in minutes based on the capacity registered in the system. But it also assumes, with the same confidence as always, that the registered capacity matches the hub's real capacity, which is currently missing two docks under renovation. What separates the expensive salesperson from the cheap one is no longer building the answer: it's knowing whether the operation can back up the promise before you close the deal.
Tuesday, the SLA dashboard turns red across an entire hub, deliveries running late since ten in the morning. Before, that meant two hours calling every driver, cross-referencing the route spreadsheet with the tracking app, trying to find the pattern. Today AI already delivers, in minutes, the list of delayed routes, grouped by likely cause: traffic, vehicle, volume above the shift's capacity. The problem is it points to "volume above capacity" with the same air of certainty as when it points to "traffic", even without knowing you already reduced the fleet at that hub two weeks ago because of maintenance. The question that separates the expensive operations professional from the cheap one is no longer "can you consolidate the dashboard?". It's: "do you know which cause is real for your hub, not just the most common one for no hub in particular?". AI points to the pattern; the real root cause is still yours.
A warehouse safety audit flags a temperature risk in the cold chamber, and AI already drafts the opinion in minutes from the sensor logs. But it also assumes, with the same air of certainty, that a one off variation is equipment failure, without knowing it was actually the door left open for the cleaning shift. What separates the expensive compliance professional from the cheap one is no longer reading the log: it's knowing which variation is a real risk.
The routing system went down for twenty minutes last night, and AI already drafts the postmortem in minutes from the service logs. But it also points to, with the same confidence as always, a cause that "seems plausible" for any similar stack, without having cross-checked that the call spike came from a reprocessing job nobody scheduled for that time. What separates the expensive technology professional from the cheap one is no longer writing the postmortem: it's finding the real root cause.
Research with app drivers points to friction in the delivery confirmation flow, and AI already summarizes the interviews in minutes. But it also assumes, with the same confidence as always, that the problem is the screen design, without knowing that half the drivers have no signal exactly in the area where confirmation gets stuck. What separates the expensive designer from the cheap one is no longer organizing the research: it's knowing where the friction really comes from.
Quarterly planning meeting, and AI already cross-references order growth with installed capacity in minutes, something that used to take days of spreadsheet work. But it also assumes, with the same air of certainty, a warehouse utilization rate that's a market average, not your actual distribution center's. What separates the expensive strategist from the cheap one is no longer building the projection: it's knowing which capacity is really yours.
Let me start with a truth that bothers anyone who works in operations. A good chunk of what used to take up your day, running the dashboard, building the schedule, cross-referencing the shift spreadsheet, writing the handoff summary, AI now does in minutes. Think about it: does that scare you or free you? The right answer depends on you understanding which part of your work became a commodity and which part became worth more. This lesson settles that account.
The core idea of this lesson. AI crashes the price of the mechanical part of operations: route calculation, schedule building, the first read of the dashboard, the draft of the shift report. That's good news, because more of you is left over for the part AI doesn't do: understanding the root cause, deciding priority when two resources compete for the same window, redesigning the process. But operations has a catch other areas don't feel the same way: if the process underneath is broken, giving it AI speed doesn't fix anything, it just speeds up the broken part. This track teaches you to operate in this new game.
01What AI commoditizes (and why that's good news)
Let's call it what it is. These things, which used to take up your hours every shift, AI already does fast and cheap:
- Route calculation and schedule building. Optimizing in seconds what used to be trial and error in a spreadsheet.
- The first read of the dashboard. "What jumped out this shift? Which SLA blew up, which hub is out of the curve?"
- The consolidation of the handoff report. Gathering what happened during the shift into one page, without you having to dig through system after system.
- The draft of the communication. The email to the customer about the delay, the internal notice to the next shift.
The big issue here is simple: when mechanical work gets cheap, it stops being your edge. The supervisor who only knew how to run the dashboard and build the schedule loses value. That sounds harsh, but there's another side, and it's the good side.
What shifts in your shift:
Here's the good news: the time you used to spend running the dashboard and building the schedule doesn't vanish, it shifts to the part that's worth more. Whoever understands this stops fearing AI and starts using it to climb altitude. Fair enough?
02What's still yours (and got more expensive)
Here's the part AI doesn't do for you, and precisely because of that, it's now worth more:
- The root cause. The dashboard says the SLA blew up. Why? Was it traffic, was it a vehicle, was it volume above the shift's capacity, was it a badly done handoff between two teams? AI points to the symptom; finding the cause is reading the real operation, and it's yours.
- The priority in conflict. Two orders, only one dock. Two routes, only one driver. Deciding who waits is yours, because that depends on context AI doesn't have: which customer is strategic, which delay costs more today.
- The redesign, not just the acceleration. Running the broken process faster isn't a solution, it's the next topic in this lesson. Redesigning the process so it stops breaking is your decision.
- The communication to the team that executes. Translating "what changes tomorrow" into a clear instruction for the next shift. That's leadership, and leadership doesn't commoditize.
Notice that all of this is judgment, not calculation. And judgment is exactly what gets expensive when calculation gets cheap. From here on, your value lives much more in "why did this happen and what do I change" than in "I managed to run the dashboard".
03Operations' deadly catch: automating a broken process just speeds up the broken part
Now the part I can't let you forget, because in operations it costs dearly in a different way than in finance. In finance, the risk is AI making up a number. In operations, the risk is you giving AI speed to a process that was already sick underneath, and the sickness simply running faster.
Think of a handoff between shifts that never had a clear owner: information always leaked a bit from one shift to the next, and someone always chased it down afterward. If you use AI only to run that handoff faster, without changing the design, you haven't fixed anything. You've just made the leak happen at higher volume, with more confidence that "it's automated, it must be right". A hidden bottleneck that today affects five orders a day can, once accelerated, affect fifty, because nobody stopped to check whether the process itself made sense.
That's why the golden rule of this track, which you'll see repeated in every choreography: before you speed up a process with AI, you audit whether the process itself is healthy. It's not distrust of the tool, it's the hygiene of whoever designs operations. AI accelerates what you already have; it doesn't fix a sick process, it just reveals the sickness faster.
04The map of this track
This lesson was the framing. The rest of the module is hands-on, installing systems into your real work, one at a time:
- Connecting AI to the operation's shop floor: system, shift spreadsheet, and sensor, each with its own path.
- The exception that resolves itself, with a human gate before any action.
- Demand forecasting with your feet on the ground, auditable instead of a confident guess.
- The living SOP, the procedure that updates itself with reality.
- Routing and allocation: optimization that respects your process's real constraint.
- The audit of the metric, for when the dashboard lies with conviction.
- And, at the end, your operations OS: the library of prompts, models, and agents that keeps working for you.
Each one takes a task you already do and redesigns it. By the end, you won't have learned about AI, you'll have installed AI into your way of operating. Shall we?
Do it now
Take an operational deliverable you did this week, your real task or another one (a shift closed, a route resolved, an SLA incident handled). Grab a sheet of paper and split it into two columns:
- Mechanical (what AI would do for you): running the dashboard, building the schedule, calculating the route, drafting the shift report.
- Judgment (what's still yours): the root cause you identified, the priority you decided between two conflicting resources, what you changed in the process, how you communicated it to the team.
Now look at the proportion. How much of your time went to the left column? That's exactly the time this track is going to give you back, for you to spend on the right column, which is the one that pays your salary.
Practice
1. In the new game of operations, what loses the most value as a professional differentiator?
2. Why is 'automating a broken process' operations' specific deadly catch?
For the board
On what got cheaprunning the dashboard, building the rota, writing the handoff. Still necessary, no longer a differentiator.
On what got expensivereading the operation, deciding under real constraints, and answering for what was delivered.
On the fatal catchhere the risk is not an invented number. It is the hidden bottleneck running faster and hitting more people.
Thanks for the feedback. It helps sharpen the next lesson.